CN112230508A

CN112230508A - Optical proximity correction method

Info

Publication number: CN112230508A
Application number: CN202011187204.2A
Authority: CN
Inventors: 顾婷婷; 张�浩; 陈翰; 宋康
Original assignee: Shanghai Huali Microelectronics Corp
Current assignee: Shanghai Huali Microelectronics Corp
Priority date: 2020-10-30
Filing date: 2020-10-30
Publication date: 2021-01-15

Abstract

The invention provides an optical proximity correction method, which comprises the following steps: importing an OPC graph into OPC software, wherein the OPC graph is provided with a plurality of specific graphs; selecting any specific graph in the OPC graphs, carrying out optical simulation on the specific graph to obtain a simulated graph, overlapping the simulated graph and the specific graph, and respectively cutting two adjacent edges of the specific graph into n sections, wherein n is more than or equal to 3; selecting one or two of the n sections as an operation section, and respectively calculating the minimum distance between each operation section and the simulation graph; and moving the corresponding operation section by a unit movement amount according to the minimum distance, performing optical simulation once every time the operation section is moved so as to update the simulation graph, and recalculating the minimum distance between each operation section and the simulation graph after updating the simulation graph every time. The invention solves the problem that the optical proximity correction of the OPC graph is influenced by grid segmentation to cause correction errors of the OPC graph in the prior art.

Description

Optical proximity correction method

Technical Field

The invention relates to the technical field of semiconductors, in particular to an optical proximity correction method.

Background

With the shrinking of the size of a semiconductor device, the requirement on the prior art is higher and higher, the requirement on the consistency of the critical dimension of the device is also higher and more rigorous, and particularly, the graphic design of a memory chip has the characteristics of high repeatability, the critical dimension of the graphic is close to the lower limit of the design dimension of the product node, and even is tighter than the lower limit of the dimension and the like. In the actual imaging process from the designed OPC pattern of the memory chip to the pattern, a certain error range is bound to exist, so that the OPC pattern of the memory chip needs to be corrected.

When OPC software is used for correcting OPC graphs, in order to accelerate processing speed, the OPC software rasterizes the OPC graphs, then corrects the sub-modules of the rasterized OPC graphs in a synchronous mode, and then integrates the sub-modules of the OPC graphs into the corrected OPC graphs. After the OPC graph is rasterized, part of graphs in the OPC graph are divided into different grids for processing, after the OPC graph is optically simulated, when OPC software processes the different grids, random sampling is carried out from the different grids, and the minimum distance from a sampling point to the OPC graph is calculated to obtain the correction movement amount in the grids; however, sampling points in one grid are different, correction movement amounts in the grid obtained through calculation are different, random sampling causes different correction movement amounts, and finally correction errors occur in the corrected OPC patterns, which may cause problems of insufficient coverage areas of upper and lower through holes after imaging or short circuits caused by connection of adjacent patterns, and finally cause low stability of the memory chip and uncontrollable risk points.

Disclosure of Invention

The invention aims to provide an optical proximity correction method to solve the problem that optical proximity correction of an OPC graph in the prior art is influenced by grid segmentation so that correction errors exist in the OPC graph.

In order to achieve the above object, the present invention provides an optical proximity correction method, including:

step S1: importing an OPC (optical proximity correction) graph into OPC software, wherein the OPC graph comprises a plurality of specific graphs, and executing step S2;

step S2: selecting any specific graph in the OPC graphs, carrying out optical simulation on the specific graph to obtain a simulated graph, overlapping the simulated graph and the specific graph, respectively cutting two adjacent edges of the specific graph into n sections, wherein n is more than or equal to 3, and executing step S3;

step S3: selecting one or two of the n sections as an operation section, respectively calculating the minimum distance between each operation section and the simulation graph, and executing the step S4;

step S4: moving the corresponding operation segment by a unit movement amount according to the minimum distance, performing optical simulation once per movement to update the simulation graph, recalculating the minimum distance between each operation segment and the simulation graph after updating the simulation graph each time, and executing step S5;

step S5: judging the size of each minimum distance and a threshold value, and executing step S4 when any minimum distance is greater than the threshold value; when each of the minimum distances is smaller than the threshold, performing step S6;

step S6: and correcting all specific graphs in the OPC graphs at one time by using the sum of the moving amount of each operation segment, and updating the OPC graphs.

Optionally, in step S4, when the corresponding operation segment is moved by the unit movement amount according to the minimum distance, all the operation segments are moved synchronously, or any one of the operation segments is moved independently.

Optionally, in step S4, the operation segments with the minimum distance smaller than the threshold are not moved.

Optionally, when n is an odd number, selecting a central section of the n sections as the operation section; and when n is an even number, selecting two central sections of the n sections as the operation sections.

Optionally, the OPC pattern corresponds to a pattern of a metal layer in the memory chip.

Optionally, the specific figure is a rectangle.

Optionally, the side length of the rectangle is 2 to 3 times of the minimum size of the metal layer design rule.

Optionally, the minimum size of the metal layer design rule is 75 nm to 100 nm.

Optionally, the lengths of the operation segments are L1 and L2, respectively, the lengths of the other segments except the operation segment in the n segments are equal and are L3, and L3 is equal to 0.7 times of the minimum size of the metal layer design rule.

Optionally, L1 is not less than L3 and not more than 2 xL 3, and L3 is not less than L2 and not more than 2 xL 3.

Optionally, the unit movement amount is 2 nm to 5 nm.

Alternatively, after step S6 is executed, step S7 is also executed,

step S7: rasterizing the updated OPC graph, and correcting other graphs except all the specific graphs in the OPC graph to obtain a target corrected graph of the OPC graph.

In the optical proximity correction method provided by the invention, any specific graph in the OPC graphs is selected, and the specific graph is subjected to optical simulation to obtain a simulated graph; respectively cutting two adjacent edges of the specific graph into n sections; selecting one or two of the n sections as an operation section, and respectively calculating the minimum distance between each operation section and the simulation graph; moving the corresponding operation section by a unit movement amount according to the minimum distance, performing optical simulation once every time the operation section is moved so as to update the simulation graph, and recalculating the minimum distance between each operation section and the simulation graph after updating the simulation graph every time; judging the size of each minimum distance and a threshold, and moving the operation section when any minimum distance is greater than the threshold; and when each minimum distance is smaller than the threshold value, finishing the movement to obtain the sum of the movement amount of each operation section, and correcting all the specific patterns in the OPC patterns at one time by the sum of the movement amount of each operation section. According to the invention, all the specific graphs are corrected at one time, and the specific graphs do not need to be corrected after being rasterized, so that correction errors caused by correction after being rasterized are avoided, the correction consistency of the specific graphs is improved, and the stability of the memory chip is finally improved.

Drawings

FIG. 1 is a diagram of a prior art memory chip with a corrected optical simulation contrast;

FIG. 2 is a flow chart of a method provided by an embodiment of the present invention;

FIG. 3 is a schematic diagram illustrating an OPC pattern and a correction thereof for a metal layer of a specific pattern in a memory chip according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a metal layer OPC pattern of a specific pattern in a memory chip, a correction schematic diagram and an optical simulation schematic diagram thereof according to an embodiment of the invention;

wherein the reference numerals are:

101-target simulation graph; 102-rasterizing the corrected simulation graph; 103-target correction pattern; 104-rasterizing the corrected pattern; 301-specific graphics; 302-a corrected pattern of the specific pattern; 401-target correction pattern of specific pattern; 402-target simulation graphics of a specific graphic; 403-upper level via holes; 404-lower level vias.

Detailed Description

Referring to fig. 1, in the prior art, OPC software is used to correct an OPC pattern of a memory chip, and after rasterizing the OPC pattern, the corrected pattern is obtained by correction. FIG. 1 shows a specific pattern in a memory chip, which is modified by the prior art to obtain a rasterized modified pattern 104, and the rasterized modified pattern 104 is optically simulated to obtain a rasterized modified simulated pattern 102; however, in fig. 1, the target simulation pattern 101 is the final pattern to be obtained by the corrected optical simulation, and compared with the target simulation pattern 101, it can be seen that the rasterized simulation pattern 102 has a slight deviation, because the rasterized correction pattern 104 has a correction error, the rasterized simulation pattern 102 has a deviation, and the target correction pattern 103 is the more accurate correction pattern to be obtained.

The following describes in more detail embodiments of the present invention with reference to the schematic drawings. The advantages and features of the present invention will become more apparent from the following description. It is to be noted that the drawings are in a very simplified form and are not to precise scale, which is merely for the purpose of facilitating and distinctly claiming the embodiments of the present invention.

Fig. 2 is a flowchart of a method provided in this embodiment, fig. 3 is a schematic diagram of a metal layer OPC pattern of a specific pattern in a memory chip provided in this embodiment and a correction diagram thereof, and fig. 4 is a schematic diagram of a metal layer OPC pattern of a specific pattern in a memory chip provided in this embodiment, a correction diagram thereof and an optical simulation diagram thereof. The invention provides an optical proximity correction method, which is used for solving the problem that the optical proximity correction of an OPC graph in the prior art is influenced by grid segmentation so that the OPC graph has correction errors. Please refer to fig. 2, which includes:

step S1: importing an OPC pattern into OPC software, wherein the OPC pattern has a plurality of specific patterns, and executing step S2;

step S3: selecting one or two of the n sections as operation sections, respectively calculating the minimum distance between each operation section and the simulation graph, and executing the step S4;

step S4: moving the corresponding operation segment by a unit movement amount according to the minimum distance, performing optical simulation once per movement to update the simulation pattern, recalculating the minimum distance between each operation segment and the simulation pattern after updating the simulation pattern each time, and executing step S5;

step S5: judging the size of each minimum distance and a threshold value, and executing the step S4 when any minimum distance is larger than the threshold value; when each minimum distance is smaller than the threshold, performing step S6;

step S6: all specific patterns in the OPC pattern are once corrected by using the sum of the moving amounts of each operation segment, and the OPC pattern is updated.

The optical proximity correction method of the present invention is described in more detail below with reference to the accompanying drawings, in which preferred embodiments of the invention are shown.

Step S1 is executed: the OPC image is imported into OPC software, and the OPC image has a plurality of specific images, and step S2 is executed.

Specifically, in this embodiment, the OPC pattern corresponds to a metal layer pattern in the memory chip, specifically, the OPC pattern is imported into OPC software, and the OPC software can support simulation of the OPC pattern and calculate to obtain a corrected OPC pattern. The OPC pattern is provided with a plurality of specific patterns, the specific patterns are metal layer connecting patterns in metal layer patterns in the memory chip, the metal layer connecting patterns usually need to meet the requirement of covering an upper through hole layer and a lower through hole layer, because the metal layer connecting patterns are used for connecting a grid and the through hole layers to form an electric path, the specific patterns are very important, if the specific patterns are divided into different grids after rasterization, correction deviation of the specific patterns is easy to occur, the electric connection problem is caused, and finally the stability of the memory chip is low. Therefore, in the present embodiment, such a specific pattern is selected for correction processing.

Step S2 is executed: selecting any specific graph in the OPC graphs, carrying out optical simulation on the specific graph to obtain a simulated graph, overlapping the simulated graph and the specific graph, respectively cutting two adjacent edges of the specific graph into n sections, wherein n is more than or equal to 3, and executing step S3.

Specifically, any specific graph in the OPC graphs is selected, the selected specific graph is subjected to optical simulation to obtain a simulated graph of the specific graph, and the simulated graph is overlapped with the specific graph. In the present embodiment, the specific pattern is a rectangle, but is not limited to this type of shape, and may be other shapes. Since the specific pattern is rectangular, the size of the simulated pattern obtained by the optical simulation is generally smaller than that of the specific pattern, and the side of the simulated pattern does not overlap with the side of the specific pattern.

Furthermore, because the specific graph in the embodiment is rectangular, and two opposite sides of the rectangle are equal, only two adjacent sides of the specific graph need to be cut into n sections, wherein n is greater than or equal to 3, but the number of the cut sections needs the length of the two adjacent sides of the specific graph to meet the cutting requirement, namely the length of each cut section needs to be greater than the minimum process requirement; in this embodiment, the length of each cut segment needs to be greater than 50 nm, where 50 nm is the current minimum process requirement, and if the cut segment needs to be 4 segments, it needs to have both sides of the specific pattern greater than 200 nm.

In this embodiment, the specific pattern is a rectangle, the side length of the rectangle is 2 to 3 times of the minimum size of the metal layer design rule, and the minimum size of the metal layer design rule is the minimum size of the current product.

Step S3 is executed: one or two of the n segments are selected as the operation segments, the minimum distance between each operation segment and the simulation graph is calculated, and step S4 is executed.

Specifically, after two adjacent sides of the specific pattern are respectively cut into n sections, one or two of the n sections are selected as the operation section. When n is an odd number, selecting a central section of the n sections as an operation section; when n is an even number, the central two sections of the n sections are selected as operation sections. The operation segment is a core segment for calculating the movement amount, and the central segment is generally selected to move so that several edges of the simulation pattern can substantially coincide with the edges of the specific pattern. Since both adjacent sides of the specific pattern are cut into n segments, both adjacent sides have operation segments, and then the minimum distance between each operation segment and the simulated pattern obtained in step S2 is calculated, which is the deviation value between the simulated pattern and the specific pattern.

Step S4 is executed: the corresponding operation segment is moved by a unit movement amount according to the minimum distance, optical simulation is performed once per movement to update the simulation pattern, the minimum distance between each operation segment and the simulation pattern is recalculated after updating the simulation pattern each time, and step S5 is performed.

Specifically, the minimum distance between each operation segment and the simulation pattern is obtained in step S3, and since the simulation pattern obtained by optically simulating the specific pattern under correction generally has a variation, the corresponding operation segment needs to be moved in accordance with a set unit movement amount, which is, in the present embodiment, in a range of 2 nm to 5 nm, but is not limited thereto. The optical simulation is performed once for each movement of the operation segment to update the simulation pattern, and the minimum distance between each operation segment and the simulation pattern is recalculated after each update of the simulation pattern.

Further, when the corresponding operation segment is moved by a unit movement amount according to the minimum distance, all the operation segments can be synchronously moved, wherein all the operation segments are each operation segment on two adjacent sides of the specific graph; or any operation segment is moved independently, namely only one operation segment in two adjacent edges of the specific graph is selected for moving.

Step S5 is executed: judging the size of each minimum distance and a threshold value, and executing the step S4 when any minimum distance is larger than the threshold value; when each minimum distance is smaller than the threshold value, step S6 is performed.

Specifically, after recalculating the minimum distance between each operation segment and the simulation pattern in step S4, a determination is made based on a set threshold, and it is determined whether or not the minimum distance between each operation segment and the simulation pattern matches the set threshold, which is the distance between the simulation pattern and the operation segment in the specific pattern, which reflects whether or not the simulation pattern substantially overlaps the edge of the specific pattern. In this embodiment, the threshold value ranges from 0.2 nm to 2 nm, but is not limited thereto, and may also range from other values, and the size of the threshold value depends on the specific requirements.

When any minimum distance is greater than the threshold value, step S4 is executed, the corresponding operation segment is moved by a unit movement amount according to the minimum distance, optical simulation is performed once per movement to update the simulation graph, after the simulation graph is updated each time, the minimum distance between each operation segment and the simulation graph is recalculated, so that after a plurality of movements, each minimum distance is required to be less than the threshold value. In step S4, the operation segments with the minimum distance smaller than the threshold are not moved, and only the operation segments with the minimum distance larger than the threshold are moved. And when each minimum distance is smaller than the threshold value, obtaining a target correction graph of the specific graph.

Step S6 is executed: all specific patterns in the OPC pattern are once corrected by using the sum of the moving amounts of each operation segment, and the OPC pattern is updated.

Specifically, in step S5, after each of the operation segments has moved, each of the minimum distances is smaller than the threshold, and the updated simulated graph substantially coincides with the edge of the specific graph, which is as expected. The sum of the movement amounts of each operation section is obtained by moving each operation section several times, and the sum of the movement amounts of each operation section is the target movement amount of each operation section. Then, all the specific patterns in the OPC pattern are corrected at once by using the target movement amount of each operation segment, and the OPC pattern is updated. And correcting all the specific graphs at one time without moving for many times, so that the obtained target corrected graphs of all the specific graphs cannot be influenced by grid segmentation to generate movement errors.

Further, after performing step S6, step S7 is also performed: and rasterizing the updated OPC graph, and correcting other graphs except all the specific graphs in the OPC graph to obtain a target corrected graph of the OPC graph.

Specifically, the OPC pattern updated in step S6 is rasterized, all the specific patterns in the updated OPC pattern are already corrected, and then other patterns except all the specific patterns in the OPC pattern are corrected to obtain the target corrected pattern of the OPC pattern. Since the OPC software sets the priority of correction and the OPC software sets the priority of correction, the rasterized correction does not re-correct all specific patterns that have been corrected, and the target corrected patterns of the OPC patterns are obtained after the other pattern corrections are completed.

Referring to fig. 3, in the present embodiment, two adjacent sides of the specific pattern 301 are respectively cut into 3 segments, the two adjacent sides are respectively designated as AD and AG, the AD is cut into an AB segment, a BC segment and a CD segment, and the AG is cut into an AE segment, an EF segment and an FG segment; the BC section and the EF section are respectively operation sections on two adjacent sides of the specific graph 301, and the lengths of the AB section, the CD section, the AE section and the FG section after cutting are the same. Since the specific pattern is rectangular, the simulation pattern after the optical simulation generally requires the shape of four corners to be the same, and therefore, the lengths of the AB segment, the CD segment, the AE segment, and the FG segment after the cutting are the same, but not limited thereto, and the lengths of the AB segment and the CD segment, and the lengths of the AE segment and the FG segment may be the same, but the AB segment and the AE segment are different. Setting the lengths of the AB section, the CD section, the AE section and the FG section as L3, the length of the BC section as L1 and the length of the EF section as L2, wherein the value of L3 is about 0.7 times of the minimum size of the metal layer design rule, 0.7 is a segmentation and segmentation coefficient, and is a preferred value of the invention, not limited to the coefficient value, but also can be other reasonable values; the values of L1 and L2 can be different, the value range of L1 is L3-L1-2 xL 3, the value range of L2 is L3-L2-2 xL 3, the BC section and the EF section are respectively the central sections of two adjacent sides of the specific graph, and the value ranges of L1 and L2 are the preferred value ranges of this embodiment and can also be the ranges of other reasonable parameter values.

And before the BC section and the EF section are not moved, carrying out optical simulation to obtain a simulation graph, respectively calculating the minimum distance between the BC section and the EF section and the simulation graph, and moving the BC section and the EF section according to whether the minimum distance is smaller than a threshold value. Assuming that the corresponding value of the BC segment in the longitudinal axis direction is zero, the above value in the longitudinal axis direction of the BC segment is positive, the below value in the longitudinal axis direction of the BC segment is negative, and the minimum distance mentioned above is a positive value, for the convenience of the following explanation of the moving process, the following measured minimum distance is assumed to have a positive or negative fraction in the longitudinal axis direction, and the actual situation is specifically set. The moving process is as follows: when the minimum distance between the BC section and the simulation graph is larger than 0 and larger than a threshold value, moving the BC section downwards along the direction of the longitudinal axis; when the minimum distance between the BC section and the simulation graph is larger than 0 and smaller than a threshold value, the BC section does not need to move any more; when the minimum distance between the BC section and the simulation graph is smaller than 0 and smaller than a threshold value, moving the BC section upwards along the longitudinal axis direction; when the minimum distance between the BC segment and the simulation graph is smaller than 0 and larger than the threshold value, the BC segment does not need to move any more.

Assuming that the corresponding value of the EF section is zero in the horizontal direction, a negative value is on the right side of the EF section in the horizontal direction, a positive value is on the left side of the EF section in the horizontal direction, and the minimum distance mentioned above is a positive value, for the convenience of the following explanation of the moving process, the following measured minimum distance is assumed to have a positive/negative fraction in the horizontal direction, and the actual situation is specifically set. The moving process is as follows: when the minimum distance between the EF section and the simulation graph is greater than 0 and greater than a threshold value, the EF section is moved to the right along the horizontal direction; when the minimum distance between the EF section and the simulation graph is larger than 0 and smaller than the threshold value, the EF section does not need to move any more; when the minimum distance between the EF section and the simulation graph is smaller than 0 and smaller than a threshold value, moving the EF section to the left along the horizontal direction; when the minimum distance between the EF section and the simulation graph is smaller than 0 and larger than the threshold value, the EF section does not need to move any more.

The unit movement amount is set according to specific conditions, optical simulation is carried out once after the operation section moves once, the minimum distances between the BC section and the EF section and the simulation graph are respectively calculated, and then the operation section continues to move according to the minimum distances. The number of shift cycles is generally set, and is determined as a practical matter, and in the present embodiment, the number of shifts is 8 to 15. In addition, the movement of the EF section and the BC section is optimally performed at the same time, and the EF section and the BC section can be moved separately, the EF section moves left and right, and the BC section moves up and down.

The target movement amount of the BC-stage and the target movement amount of the EF-stage are obtained by the sum of the unit movement amounts obtained by the plurality of movements, and the movement amount of the BC-stage is assumed to be x, and the movement amount of the EF-stage is assumed to be y. The dotted line graph in the figure is the corrected graph 302 of the specific graph, the corrected graph 302 of the specific graph needs to meet the requirement of the subsequent photomask processing, the photomask factory needs to perform photomask rule check on the corrected graph during photomask processing, h in the figure is a photomask rule check limit value given by the photomask factory, and the corrected graph needs to meet the size of the h value, in the embodiment, the value range of h is 20 nanometers to 50 nanometers, and can also be other rule limit values. Wherein the value of x is influenced by the parameters L1, L3 and h, and the value of y is influenced by the parameters L1, L2 and h. The linear relation between the moving amounts of the AB section and the CD section and the BC section is obtained through a plurality of experiments, and the moving amounts of the AB section and the CD section can be controlled to a certain degree by controlling the moving amount of the BC section; the movement amount of the AE section and the EG section has a linear relation with the EF section, and the movement amount of the AE section and the EG section can be controlled to a certain degree by controlling the movement amount of the EF section. Through multiple batch experimental simulation, the optimal values of x and y in a specific graph can be determined to be x_fAnd y_fAnd obtaining the optimal target movement amount of the two adjacent edge operation segments of the specific graph, wherein the target movement amount of the other two edges is the same as that of the opposite edge.

Referring to fig. 4, the target correction pattern 401 of the specific pattern can be obtained by the optimal target movement amount of the two adjacent side operation segments of the specific pattern 301, and the target correction pattern 401 of the specific pattern is optically simulated to obtain the target simulation pattern 402 of the specific pattern, wherein four sides of the target simulation pattern 402 of the specific pattern substantially coincide with four sides of the specific pattern 301, and the target simulation pattern 402 of the specific pattern completely covers the upper layer through hole 403 and the lower layer through hole 404, so that the specific pattern 301 can be better corrected, and the correction error can be reduced.

In summary, the present invention provides an optical proximity correction method, which performs optical simulation on a specific pattern by selecting any specific pattern in OPC patterns to obtain a simulated pattern; respectively cutting two adjacent edges of the specific graph into n sections; selecting one or two of the n sections as an operation section, and respectively calculating the minimum distance between each operation section and the simulation graph; moving the corresponding operation section by a unit movement amount according to the minimum distance, performing optical simulation once every time the operation section is moved so as to update the simulation graph, and recalculating the minimum distance between each operation section and the simulation graph after updating the simulation graph every time; judging the size of each minimum distance and a threshold, and moving the operation section when any minimum distance is greater than the threshold; and when each minimum distance is smaller than the threshold value, finishing the movement to obtain the sum of the movement amount of each operation section, and correcting all the specific patterns in the OPC patterns at one time by the sum of the movement amount of each operation section. According to the invention, all the specific graphs are corrected at one time, and the specific graphs do not need to be corrected after being rasterized, so that correction errors caused by correction after being rasterized are avoided, the correction consistency of the specific graphs is improved, and the stability of the memory chip is finally improved.

The above description is only a preferred embodiment of the present invention, and does not limit the present invention in any way. It will be understood by those skilled in the art that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

1. An optical proximity correction method, comprising:

2. The optical proximity correction method of claim 1, wherein in step S4, when the corresponding operation segment is moved by the unit movement amount according to the minimum distance, all the operation segments are moved simultaneously or any one of the operation segments is moved individually.

3. The optical proximity correction method of claim 2, wherein in step S4, the operation segment whose minimum distance is less than the threshold value is not moved.

4. The optical proximity correction method according to claim 1, wherein when n is an odd number, a central one of the n segments is selected as the operation segment; and when n is an even number, selecting two central sections of the n sections as the operation sections.

5. The optical proximity correction method of claim 1, wherein the OPC pattern corresponds to a pattern of a metal layer in the memory chip.

6. The optical proximity correction method of claim 5, wherein the specific pattern is a rectangle.

7. The method of claim 6, wherein the rectangle has a side length of 2 to 3 times a minimum dimension of the metal layer design rule.

8. The method of claim 7, wherein the minimum dimension of the metal layer design rule is 75 nm to 100 nm.

9. The method of claim 8, wherein the lengths of the operation segments are L1 and L2, respectively, the lengths of the other ones of the n segments excluding the operation segment are equal and are L3, and L3 is equal to 0.7 times the minimum size of the metal layer design rule.

10. The optical proximity correction method of claim 9, wherein L3 ≦ L1 ≦ 2 xl 3 and L3 ≦ L2 ≦ 2 xl 3.

11. The optical proximity correction method of claim 1, wherein the unit shift amount is 2 nm to 5 nm.

12. The optical proximity correction method of claim 1, wherein after the step S6 is executed, a step S7 is further executed,